Abstract

The evolution of the bird palate remains confusing after 150 years of discussion. A pterygoid-vomer contact is considered typical of Palaeognathae, and a pterygoid-palatine contact, of Neognathae. Embryological studies show that in Galloanserae the palatines ossify before the vomer (the opposite of Palaeognathae) and a pterygoid-vomer contact is never established. This pattern is widely assumed to represent Neognathae, but contrasts with reports of a pterygoid-vomer contact in juveniles of several Neoaves, the other main lineage of Neognathae. In several Neoaves, an offset region of the anterior pterygoids (the ―hemipterygoid‖) contacts the vomer in juveniles and then breaks off the main body in the adult, fusing with the palatines. We generated our own series of embryological development for several orders of birds and re-analyzed juvenile (post-hatching) and adults in museum collections. We found that in many Neoaves with a large vomer, a connection with the hemipterygoid may persist in the adult (Apodidae, Megalaimidae, Trochiilidae, Indicatoridae and Jinginae) as previously reported, but this is often hard to establish because of seamless fusion (sinostosis). The hemipterygoid is also present in birds with a reduced or absent vomer, but only ever contacts the palatines. In Palaeognathae, the hemipterygoid contacts the vomer during embryonic development. This is not the case in Neoaves with a large vomer: The vomers ossify after the palatines, and only come close to the pterygoids at much later, near-hatching stages. Their contact in Neoaves was only observed at post-hatching stages. We suggest the Neoavian pattern represents early Neognathae: The pattern of Galloanserae (where the vomers are smaller, ossify even later, and never contact the pterygoids) could be derived, and thus misleading as a characterization of Neognathae.